In synizesis two vowels that would normally be pronounced separately are pronounced as one without any change in spelling. This happens sometimes in Latin poetry and it can be recognized from the metric — sometimes proper scansion is impossible without synizesis.

Synizesis seems to be somewhat rare. How rare exactly was it in Latin poetry of Augustan era? I am looking for numerical answers for the frequency of synizesis, answering either of the following questions:

  • How many times does synizesis occur in a verse on average?
  • In what percentage of the places where synizesis is possible does it actually take place?

I believe the frequency of synizesis varied between poets and eras, and therefore I restricted my attention to Augustan poets (Vergilius, Horatius, Ovidius). It may also vary significantly between the works of individual authors. If such synizesis data is available for the main works of all these authors, please give a couple of examples (like data for Aeneis or Metamorphoses) in addition to instructions for finding more information.

(I chose to make this question slightly open-ended in order to make it more easily answerable. If you feel that this question should be more specific (only about Aeneid, for example), let me know and I will reformulate it.)

1 Answer 1


This is a simple-minded study of a very specific case of synizesis to get a concrete numerical example. I am looking at the letters -eo- (next to each other in the same word) in the first two books of the Aeneid. A corpus search gives 281 hits for the whole Aeneid, so any results will have some significance but the amount of work is not utterly unreasonable.


I will divide instances of -eo- in three categories based on scansion:

  • A: It has to scan as short+long.
  • B: It has to scan as long or short+short.
  • C: It has to scan as long.
  • D: Other.

Category A means no synizesis, B means either synizesis or (iambic or other) shortening, and C means synizesis. D is for cases where the e belongs to ae or the o is elided or something similar interferes with the two vowels. I only want cases where the e is short and not part of a diphthong, but the o can be short or long.


Here are the findings based on the verses listed below:

Book A B C D
1 13 0 0 9
2 27 0 0 1

Category D is to be dismissed, as synizesis only becomes an option in categories B and C. But there are zero instances in the two books!

If Vergil had allowed himself to synizesize eo into a single syllable, I have to imagine that he would have done so, given the number of occurrences here. The conclusion from this amount of data is that synizesis of eo is very rare in Vergil.

Looking at different vowels and different authors and filtering the results differently will certainly yield different numbers. The frequency of synizesis can vary greatly between authors and works. I don't want to continue this specific exercise beyond the first two books, but I do know that there is a clear case of C in 6.412:

deturbat laxatque foros; simul accipit alueo

If someone has a tool that can automatically scan poetry, perhaps it could be used to identify possible cases of synizesis in a larger collection of poems. A solid collection of positives would give a firmer ground to any statistical analysis.

The verses

1.6 A:

inferretque deos Latio; genus unde Latinum

Aeolia and Aeolus on lines 52, 52, 56, 65, 75, 141 of book 1. All D.

1.133 A:

iam caelum terramque meo sine numine, uenti

1.254 A:

Olli subridens hominum sator atque deorum

1.303 A:

corda uolente deo; in primis regina quietum

1.337 A:

purpureoque alte suras uincire coturno.

1.413 A:

cernere ne quis eos neu quis contingere posset

1.416 D:

laeta suas, ubi templum illi, centumque Sabaeo

1.489 D (the E is long):

Eoasque acies et nigri Memnonis arma.

1.543 A:

at sperate deos memores fandi atque nefandi.

1.567 A:

non obtunsa adeo gestamus pectora Poeni,

1.575 A:

atque utinam rex ipse Noto compulsus eodem

1.589 A:

os umerosque deo similis; namque ipsa decoram

1.649 A:

et circumtextum croceo uelamen acantho,

1.651 D:

Pergama cum peteret inconcessosque hymenaeos,

1.671 A:

uocibus, et uereor quo se Iunonia uertant

1.711 A:

pallamque et pictum croceo uelamen acantho.

2.49 A:

quidquid id est, timeo Danaos et dona ferentis.

2.134 A:

eripui, fateor, leto me et uincula rupi,

2.159 A:

si qua tegunt, teneor patriae nec legibus ullis.

2.181 A:

arma deosque parant comites pelagoque remenso

2.263–264 A & D:

Pelidesque Neoptolemus primusque Machaon
et Menelaus et ipse doli fabricator Epeos.

2.320 A:

sacra manu uictosque deos paruumque nepotem

2.365 A:

corpora perque domos et religiosa deorum

2.371 A:

Androgeos offert nobis, socia agmina credens

2.382 A:

haud secus Androgeos uisu tremefactus abibat.

2.389 A:

mutemus clipeos Danaumque insignia nobis

2.417 A:

confligunt, Zephyrusque Notusque et laetus Eois

2.422 A:

apparent; primi clipeos mentitaque tela

2.431 A:

Iliaci cineres et flamma extrema meorum,

2.443 A:

nituntur gradibus clipeosque ad tela sinistris

2.500 A:

caede Neoptolemum geminosque in limine Atridas,

2.549 A:

degeneremque Neoptolemum narrare memento.

2.567 A:

Iamque adeo super unus eram, cum limina Vestae

2.587 A:

ultricis famam et cineres satiasse meorum

2.593 A:

continuit roseoque haec insuper addidit ore

2.613 A:

sufficit, ipse deos in Dardana suscitat arma

2.632 A:

descendo ac ducente deo flammam inter et hostis

2.671 A:

Hinc ferro accingor rursus clipeoque sinistram

2.700 A:

adfaturque deos et sanctum sidus adorat.

2.722 A:

ueste super fuluique insternor pelle leonis

2.734 A:

ardentis clipeos atque aera micantia cerno.

2.745 A:

quem non incusaui amens hominumque deorumque

2.764 A:

incensis erepta adytis, mensaeque deorum

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.